Journal of Hepatology 43 (2005) 711–719 www.elsevier.com/locate/jhep

Pathways for the regulation of body homeostasis in response to experimental

Igor Theurl1, Susanne Ludwiczek1, Philipp Eller1, Markus Seifert1, Erika Artner2, Peter Brunner3,Gu¨nter Weiss1,*

1Department of General Internal Medicine, Clinical Immunology and Infectious , Medical University, Anichstr. 35, A-6020 Innsbruck, Austria 2Institute for Medical Chemistry, Medical University of Innsbruck, A-6020 Innsbruck, Austria 3Institute of Analytical Chemistry and Radiochemistry, University of Innsbruck, Innsbruck, Austria

Background/Aims: Secondary iron overload is a frequent clinical condition found in association with multiple blood transfusions. Methods: To gain insight into adaptive changes in the expression of iron in duodenum, and spleen upon experimental iron overload we studied C57BL/6 mice receiving repetitive daily injections of iron-dextran for up to 5 days. Results: Iron initially accumulated in spleen but with subsequent increase in and expression its content in the spleen decreased while a progressive storage of iron occurred within hepatocytes which was paralleled by a significant increase in and expression. Under these conditions, iron was still absorbed from the duodenal lumen as divalent metal transporter-1 expressions were high, however, most of the absorbed iron was incorporated into duodenal ferritin, while ferroportin expression drastically decreased and iron transfer to the circulation was reduced. Conclusions: Experimental iron overload results in iron accumulation in macrophages and later in hepatocytes. In parallel, the transfer of iron from the gut to the circulation is diminished which may be referred to interference of hepcidin with ferroportin mediated iron export, thus preventing body iron accumulation. q 2005 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.

Keywords: Iron; Hemochromatosis; Iron absorption; Hepcidin; IRP

1. Introduction Secondary iron overload leads to progressive iron accumulation in the reticulo-endothelial system [1] with Secondary iron overload is a frequent clinical condition subsequent negative effects on cell mediated immune which can arise from inborn errors of haemoglobin function, since iron inhibits interferon-g inducible pathways synthesis, myelodysplastic syndromes or chemotherapy in macrophages [4]. Accordingly, subjects with secondary induced anaemia, which all require multiple blood transfu- iron overload are at a higher risk for tuberculosis and cancer sions [1,2]. In addition, an inherited form of secondary iron [3]. In being a catalyst for the formation of highly toxic overload is found in Southern Africa which has been linked radicals, iron overload leads to tissue damage and organ to a defect in duodenal iron transport together with the failure [5], which is well known in primary iron overload, consumption of iron rich traditional beer [3]. hereditary hemochromatosis, although with this condition iron is primarily deposited within parenchymal cells [6–8]. However, information of adaptive changes of body iron Received 29 November 2004; received in revised form 9 March 2005; homeostasis upon secondary or experimental iron overload accepted 17 March 2005; available online 13 June 2005 is rare. * Corresponding author. Tel.: C43 512 504 23255; fax: C43 512 504 25607. Maintenance of body iron homeostasis is mainly E-mail address: [email protected] (G. Weiss). regulated by duodenal iron absorption. Thereby, ferric

0168-8278/$30.00 q 2005 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jhep.2005.03.030 712 I. Theurl et al. / Journal of Hepatology 43 (2005) 711–719 iron is reduced at the luminal site by Animals used for experiments were all between 12 and 16 weeks of age, with an average weight of 25 g. For induction of iron overload animals were (Dcytb) [9], and ferrous iron is then transferred into the w daily intraperitoneally injected with 1 mg of iron dextran (venofer ) for up by means of the transmembrane divalent to 5 days. metal transporter-1 (DMT-1) [10,11]. At the basolateral site Mice were anaesthetized, blood was collected by orbital puncture, and ferrous iron is exported from the enterocyte to the the animals were then sacrificed. Tissue samples were frozen in liquid nitrogen or fixed in formalin for . circulation via ferroportin [12–14], and after being oxidised by the membrane bound ferroxidase [15] iron is incorporated into . In addition, up to 50% of 2.2. RNA extraction, northern blot analysis absorbed iron may be in the form, which is taken up and Real Time PCR by a yet not characterised duodenal heme receptor [16,17]. Total RNA was extracted from nitrogen frozen tissue samples using a The expression of these iron transport genes is strongly guanidinium–isothiocyanate–phenol–chloroform based procedure as pre- regulated by body iron homeostasis, since the duodenal viously described and subjected to northern blotting [43]. For hybridisation expression of DMT-1, Dcytb, ferroportin and to a lesser we generated radiolabelled cDNA probes applying the oligoprimer procedure using a 2.3 kb EcoRI insert of murine TfR1, a 1.8 kb EcoRI extent of hephaestin are increased with iron deficiency insert of murine DMT-1, a 650 bp EcoRI insert of murine ferroportin, a anaemia [9,11,15,18–22]. The presence of iron responsive 950 bp EcoRI insert of murine Dcytb,a 1.5 kb EcoRI insert of murine elements (IREs) within the untranslated regions of DMT-1 hephaestin, a 250 bp EcoRI insert of murine Hepcidin, a 900 bp EcoRI insert of murine HFE and a 1.9 kb HindIII insert of chicken beta actin of a and ferroportin mRNA suggested that the expression of PCR2.1 Vector. In parallel, quantitative PCR was carried out exactly as these may be susceptible to posttranscriptional described [43]. The following primers (a, forward primer; b, reverse 0 regulation by iron regulatory proteins (IRP) [23]. However, primer) and TaqMan probes (c, TaqMan probe) were used: muTfR1: (a) 5 - CGCTTTGGGTGCTGGTG-30,(b)50-GGGCAAG TTTCAACAGAA transcriptional regulation of DMT-1 and ferroportin GACC-30, (c) 50-CCCACACTGGACTTCGCCGCA-30, muHFE: (a) 50- expression by iron has also been anticipated [24–26]. TCATGAGAGTCGCCGTGCT-30, (b) 50-GGCTTGAGGTTTG CTCCA- 0 0 0 In terms of regulation of duodenal iron absorption [27] 3 , (c) 5 -AGCCCAGGGCCCCGTGGAT-3 , muDMT-1(IRE-form): (a) 50-CCAGCCAGTAAGTTCAAGGATCC-30,(b)50-GCGTAGCAGCT (TfR1) mediated uptake of circulating GATC TGGG-30, (c) 50-TGGCCTCGCGCCCCAACA-30, muFerropor- iron from the basolateral site of and the tin:(a)50-CTACCATTAGAAGGATTGACCAGCT-30,(b)50- 0 0 subsequent modulation of IRP activity was believed to be CAAATGTCATA ATCTGGCCGA-3 , (c) 5 -CAACATCCTGGCCCC- CATGGC-30, muHepcidin: (a) 50-TGTCTCCTGCTTCTCCTCCTTG-30, the pivotal mechanism for sensing the body’s needs for iron (b) 50-AGCTCTGTA GTCTGTCTCATCTGTTGA-30, (c) 50-CAGCCT- to the enterocyte [16,17]. The membrane bound protein GAGCAGCACCACCTATCTCC-30. muDcytb:(a)50 -TCGCGG- 0 0 0 0 HFE which is mutated in approximately 80% of subjects TGACCGGCT-3 ,(b)5-TTCCAGGTCCATGGCAGTCT-3 ,(c)5- CGTCTTCATCCAGGGCATCGCC-30, muHephaestin:(a)50-TGG- with hereditary hemochromatosis [28], modulates the AACT ATGCTCCCAAAGGAA-30, (b) 50-CTGTTTTTGCCAGACTT affinity of transferrin for TfR1 but its role in iron absorption CAGGA-30, (c) 50-CAAATCAGACTCTCAACAATGACACAGTGGCT- 0 0 0 0 remains elusive [29]. Importantly, the recent identification 3 . muHemojuvelin: (a) 5 -GGT TTC GTC GGG AGC CA-3 , (b) 5 -TGG TAG ACT TTC TGG TCA ATG CA-30, (c) 50-CGC TAC CAC CAT CCG of the liver cationic peptide hepcidin has suggested that this GAA GAT CAC TAT CAT ATT-30, muTNFa: (a) 50-TTCTATGGCC- molecule may be the principal regulator of iron absorption. CAGACCCTCA-30, (b) 50-TTGCTACGACGTGGGCTACA-30, (c) 50- 0 [30–32]. The expression of hepcidin is increased when body CTCAGATCATCTTCTCAAAATTCGA GTGACAAGC-3 , muIL6: (a) 50-GAGGATACCACTCCCAACAGACC-30, (b) 50-AAGTGCATCATC iron stores are high [32–34]. Hepcidin over-expression GTTGTTCATACA-30, (c) 50-CAGAATTGCCATTGCACAA CTCTTT resulted in reduced duodenal iron uptake and the develop- TCTCA-30. ment of anaemia [33,35] while reduced hepcidin expression causes iron overload [36–39]. This, may be traced back to a 2.3. Iron regulatory protein activity direct interaction of hepcidin with ferroportin expression, thus affecting iron transfer from the duodenal enterocyte to Protein-extracts were prepared from nitrogen frozen tissue homogen- the circulation [40,41]. The liver derived protein hemoju- ized in cytoplasmic lysisbuffer (25 mM Tris–HCl pH 7.4, 40 mM KCl, 1% Triton X-100) containing 1 mg/ml of each aprotinin, leupeptin and velin may have additive effects to those of hepcidin [42]. phenylmethylsulfonyl flouride. For gel retardation assays a 32P-labelled In the study presented here, we used a mouse model of IRE probe was prepared [44], and the analysis of RNA/protein complexes experimental iron overload to study adaptive changes of was carried out by non-denaturing gel electrophoresis and subsequent auto- radiography as described [45]. iron transport molecules in duodenum, liver and spleen in order to get new insights into regulation of iron homeostasis upon secondary iron overload. 2.4. Western blotting

For western blotting 20 mg of protein extracts, prepared as described for the gel retardation assay, were run either on a 10% (for TfR1) or a 15% (for ferritin and b-actin) SDS-polyacrylamide gel. Proteins were transferred 2. Material and methods onto a nylon membrane (Hybond-P, Amersham-Pharmacia, Vienna, Austria) and blocked in 1!TBS buffer containing 5% dry milk and 0.1% Tween (Merck, Vienna, Austria). The membrane was incubated either with 2.1. Animal care human anti-TfR1-antibody (0.5 mg/ml, Zymed, Vienna, Austria), human anti-ferritin-antibody (2 mg/ml, Dako, Vienna, Austria), murine anti-HFE- Male C57BL/6 mice were kept according to institutional and antibody (a generous gift by Dr M. Chorney [46]) or human anti-b-actin governmental guidelines on a standard rodent diet in the animal quarter (2 mg/ml, Sigma, Germany) and further processed as described [43]. b-actin at the University of Innsbruck. hybridization was used to demonstrate equal protein loading onto gels. I. Theurl et al. / Journal of Hepatology 43 (2005) 711–719 713

2.5. Immunohistochemistry and Perls’ 3. Results When investigating tissue iron content following iron Formalin fixed, paraffin embedded tissue specimens were used. High loading over time we found progressive iron accumulation temperature antigen unmasking was performed using Antigen Unmasking in liver (Table 1). In the spleen, iron content first increased Solution (Vector Laboratories, Burlingame, CA). To inactivate endogen- ous peroxidase, the tissue was incubated with Peroxidase Blocking significantly over 3 days before it descended in spite of Reagent (DAKO, USA, Carpinteria, CA) for 30 min followed by a continuing iron loading (Table 1). 30 min incubation with diluted goat normal serum using Elite Vectastain ABC Kit (Vector Laboratories) to reduce background staining. Subsequently, tissue sections were incubated at 4 8C over night with 3.1. Changes in the expression of iron genes polyclonal rabbit anti mouse antibody direct against ferroportin, DMT-1 in duodenum, liver and spleen upon experimental and Dcytb (Alpha Diagnostic international, San Antonio, USA), iron overload respectively. All subsequent steps were performed exactly as suggested by the manufacturer using Peroxidase Substrate Kit AEC (Vector Laboratories, CA). For quality control we performed immunohistochem- To get an impression on the dynamic processes occurring istry by using only the second antibody, and in addition also by using an after progressive iron loading we performed northern blots isotype-matched control for mouse IgG. In both circumstances, we did for target iron genes. In order to statistically quantify changes not get a specific staining. Iron in tissue sections was detected by Perls’ Prussian Blue staining in expression patterns between baseline and at the end following a standard protocol [47]. of cumulative iron overload RT-PCR was performed. TfR1 mRNA levels progressively decreased in the liver as estimated by RT-PCR with iron loading over time, 2.6. Quantitative measurement of duodenal iron while they significantly increased in the duodenum uptake in vivo (Table 2). In the spleen, TfR1 mRNA expression decreased over 3 days of iron administration but increased thereafter For radioactive iron uptake assays control mice and mice having received five consecutive injections of iron dextran were orally fed with (Fig. 1). Thus, TfR1 mRNA levels inversely paralleled the 100 ml of 1.05 mg/ml 59Ferric chloride (specific activity O3 mCi/mg 59Fe) changes in tissue iron content in liver and spleen (Tables 1 in 0.01 M hydrochloric acid using a gastric tube. All mice were fasted 24 h and 2). before initiation of the oral iron uptake assay. Thirty minutes after oral iron administration, mice were anaesthetized, blood was collected by orbital While no significant change in DMT-1 expression was puncture, and the animals were then sacrificed. The proximal 3 cm of the observed in the liver it decreased with prolonged iron duodenum were cut out and extensively rinsed. Radioactive iron content in overload in the spleen (Fig. 1, Table 2). In the duodenum, serum and duodenum were measured using a gamma-counter. In parallel, 59 DMT-1 and Dcytb mRNA levels initially decreased upon the incorporation of Fe into duodenal ferritin-core-complexes was investigated as follows: Duodenal slices were subjected to cytosolic induction of iron overload but then ascended with prolonged protein preparation, as described for gel retardation assay, and separated on iron substitution (Fig. 1). In contrast, Dcytb mRNA a 8% SDS-polyacrylamide gels under non-reducing conditions with concentration increased with iron loading in the liver and subsequent auto-radiography. spleen (Table 2). Ferroportin mRNA levels increased significantly in liver 2.7. Chemical iron measurements and spleen and decreased in the duodenum upon progressive iron loading (Table 2, Fig. 1), and hephaestin mRNA The concentration of iron in tissues was measured by atomic absorption expression paralleled these changes (Table 2, Fig. 1), spectrometry. Liver and spleen iron content was determined after acid digestion of tissue samples followed by iron quantification with atomic although not being significant in the spleen. absorption spectrometry as described [23]. Values are expressed in In the liver, both hepcidin and hemojuvelin mRNA levels microgram per gram wet weight. significantly increased with iron loading over time. Hepcidin levels were maximal after 3 days of iron loading 2.8. Data analysis and remained high until day 6, thus being significantly higher than at baseline (Table 2). No difference in TfR-2 Statistical analysis was carried out using SPSS statistics package and mRNA expression was observed between control and iron utilized Spearman’s rank correlation and one-sided ANOVA. overloaded mice (Table 2).

Table 1 Changes of tissue iron content upon progressive iron loading

Cd1 d3 d6 P Liver 167 (G39) 636 (G191) 1270 (G693) 2091 (G792) 0.000 Spleen 197 (G48) 357 (G27) 402 (G157) 164 (G5) 0.045

Mice were divided into four different groups (nZ8 each) receiving daily intraperitoneal injections of iron dextran for 1, 3 and 6 days plus one control group. All mice were sacrificed the same day and total tissue iron content was determined by atomic absorption in liver and spleen tissue slices. Data are expressedas means of microgram iron per gram tissueCSD. P-values annotate trends for changes in iron content over time (control to day 6) as estimated by ANOVA analysis. 714 I. Theurl et al. / Journal of Hepatology 43 (2005) 711–719

Table 2 Changes over time in liver, duodenum and spleen mRNA levels of iron metabolism and transport molecules upon experimental iron load in mice

Duodenum Liver Spleen C 6dPC 6dPC 6dP TfR1 1.14 (G0.14) 2.17 (G0.29) 0.028 2.00 (G0.31) 1.40 (G0.17) 0.036 1.77 (G0.84) 5.51 (G2.07) 0.077 HFE 0.96 (G0.04) 1.60 (G0.18) 0.031 0.97 (G0.05) 1.64 (G0.21) 0.05 1.82 (G0.84) 1.16 (G0.63) 0.118 DMT-1 0.94 (G0.08) 1.15 (G0.16) 0.624 1.03 (G0.16) 1.33 (G0.14) 0.494 1.12 (G0.26) 0.38 (G0.18) 0.056 Ferroportin 1.39 (G0.21) 0.91 (G0.12) 0.033 1.01 (G0.11) 2.75 (G0.24) 0.000 1.03 (G0.06) 3.14 (G1.03) 0.041 Dcytb 2.05 (G0.74) 0.87 (G0.24) 0.041 0.77 (G0.34) 1.58 (G0.68) 0.021 2.72 (G1.5) 4.04 (G1.8) 0.286 Heph 2.84 (G0.37) 1.59 (G0.27) 0.005 0.81 (G0.27) 1.82 (G0.94) 0.02 1.95 (G1.27) 3.29 (G2.84) 0.566 TfR2 1.51 (G0.74) 1.28 (G0.49) 0.337 Hepcidin 1.03 (G0.22) 4.90 (G0.88) 0.001 Hemojuvelin 1.07 (G0.71) 1.56 (G0.41) 0.014

Transferrin receptor (TfR1), HFE, divalent metal transporter-1 (DMT-1,) ferroportin, duodenal cytochrome b oxidase (Dcytb), Hephaestin (Heph), Hepcidin and hemojuvelin mRNA levels of eight mice in each group were quantified by RT-PCR analysis. Values for control and 6 days iron loaded mice are expressed as relative abundance normalized to 18S rRNA levels and are presented as meanCSD. P-values are calculated by oneway Anova.

Finally we determined IL-6 and TNF-a mRNA levels in carried out. In the liver, IRP binding activity decreased over spleen and liver samples by means of RT-PCR. We could time with iron loading while in the spleen IRP binding not found any significant difference in the expression of affinity initially decreased but then increased after 6 days of these cytokines between control and iron loaded animals iron overload (Fig. 2), thus paralleling the changes in iron (data not shown). TfR2 and hemojuvelin were hardly content in both tissues. In contrast, IRP binding activity in detectable by northern blots technique and their expression the duodenum increased with iron administration over time was thus only determined by quantitative PCR. (Fig. 2, Table 1).

3.2. Determination of intracellular iron availability 3.3. Modulation of organ specific protein expression by bandshift assays of critical iron genes by secondary iron overload

To estimate the effects of secondary iron overload on In keeping with mRNA data TfR1 protein expression, as intracellular iron availability in different tissues, RNA determined by western blotting, was inversely regulated in bandshift assays for determination of IRP activity were liver and duodenum (Fig. 3). In the spleen, TfR1 protein

Fig. 1. Duodenal, liver and spleen mRNA levels of iron metabolism and transport molecules in control and secondary iron loaded mice determined by northern blot analysis. Mice divided in four different groups received none (C), one, three or five daily i.p. injections of iron dextran. All mice were sacrificed the same day and duodenal, liver and spleen tissue were subjected to RNA preparation for northern blotting. One of four representative experiments is shown for TfR1, HFE, DMT-1, ferroportin, Dcytb, Hephaestin, Hepcidin and beta-actin, respectively. RNA expression in some tissues was to low for detection by means of northern blot technique. These cases are identified by n.d. for not detected. I. Theurl et al. / Journal of Hepatology 43 (2005) 711–719 715

Fig. 2. Effect of secondary iron overload on IRP binding activity in duodenum, liver and spleen. Duodenum, liver and spleen tissue from mice of the four different groups was subjected to cytosolic protein preparation for following bandshift assays. One of four representative experiments is shown. Total IRP amounts were determined by incubation with 2-mercaptoethanol. expression first decreased and then returned to baseline was higher in iron loaded mice than in controls. In contrast, expression, paralleling changes in IRP binding activity Dcytb showed no specific staining in the liver (A3, A7) and (Fig. 3). is therefore not shown in the figure. Perls’ Prussian Blue Moreover, HFE protein levels mirrored mRNA levels staining showed increased iron accumulation in iron showing maximum HFE protein expression in liver and overloaded animals (A8, C8) as compared to controls (A4, duodenum after 3–6 days of iron loading (Fig. 3). C4) in liver and spleen. However, iron accumulation in the Ferritin protein levels in liver increased with prolonged spleen was strongest in mice loaded only for 3 days (not iron challenge while no significant changes were found in shown). The staining could be referred to iron storage within duodenum. Spleen ferritin levels were elevated already after cells of the reticulo-endothelial system. Moreover, an iron 1 day and remained high thereafter (Fig. 3). In duodenum deposition in parenchymal liver cells could be found after and spleen, both H-ferritin and L-ferritin, were detectable 6 days of iron loading (Fig. 4). already in control mice, whereas in liver H-ferritin, reflecting the lower band in the western blot of mice [48], 3.4. Correlations between liver hepcidin mRNA and iron could not be detected until 6 days of iron loading (Fig. 3) transporter mRNA in different tissues In accordance with mRNA data (Fig. 1) ferroportin protein expression decreased with prolonged iron overload Spearman rank correlation technique was applied to in the duodenum (B1, 5). In the liver of control mice, study the cross-regulatory interactions between hepcidin and ferroportin (A1) was detectable in the sinusoidal borders of iron transporter mRNA expression in the tissues investi- hepatocytes and Kupffer cells at low levels. During the iron gated. In control animals, liver hepcidin mRNA levels loading phase ferroportin expression strongly increased in correlated negatively to duodenal DMT-1 mRNA Kupffer cells and vanished in hepatocytes (A5). Ferroportin (rZK0.673, PZ0.033) and positively to hepatic TfR1 in spleens of iron loaded mice (C5) was restricted to the red mRNA (rZ0.873, PZ0.000) levels. Following iron pulp, reflecting reticulo-endothelial cells. overload we found a positive association between hepcidin DMT-1 tissue expression in liver (A2, A6) and spleen and ferroportin levels in liver (rZ0.687, PZ0.000) and (C2, C6) was not significantly altered by secondary iron spleen (rZ0.461, PZ0.05) while a negative correlation overload while the apical DMT-1 staining in the duodenum was evident in duodenum (rZK0.561, PZ0.005).

Fig. 3. Effect of secondary iron overload on protein levels of iron metabolism and transport molecules in different tissues. The same tissues as used for RNA preparation were subjected to protein preparation for subsequent western blotting and determination of protein levels of TfR1, HFE, ferritin and beta-actin in duodenum, liver and spleen. One of four representative experiments is shown. 716 I. Theurl et al. / Journal of Hepatology 43 (2005) 711–719

Fig. 4. Tissue distribution of ferroportin, DMT-1, Dcytb and Perls’ Prussian Blue staining in liver, duodenum and spleen. Bright field photographs of mouse liver (panel A), duodenum (panel B) and spleen (panel C) immunohistochemistry using affinity purified anti-ferroportin (1, 5), anti-DMT-1 (2, 6), anti-Dcytb antibody (3, 7) and Perls’ Prussian Blue staining (4, 8). Control tissue (1–4) and tissue of 6 times injected (cumulative dose 15 mg iron dextran) mice (5–8) correspond to the stated numbers. A3 and A7 are not shown, as there was no specific staining with Dcytb antibody in the liver. All sections were paraffin-embedded. Antibody concentrations were 2.5 mg/ml for anti-ferroportin and anti-Dcytb and 5 mg/ml for anti-DMT-1 antibody. One of eight slides is shown each.

No significant correlations between hepcidin mRNA and iron overloaded group (Fig. 5c). Accordingly, the transfer of HFE, Dcytb or hephaestin expressions, respectively, were iron form the duodenal enterocyte to the circulation was observed at any time in the organs investigated (not shown). significantly reduced in iron overloaded species as compared to controls (Fig. 5b). 3.5. Impact of secondary iron overload on duodenal iron absorption 4. Discussion Control mice and mice after 5 consecutive days of intraperitoneal iron injection were subjected to oral feeding We herein investigated the dynamic changes in the with 59Ferric chloride using a gastric tube. The content of expression patterns of regulatory iron genes in different radioactive iron in duodenal slices was significantly higher organs following experimental iron overload. in iron overloaded than in control mice (Fig. 5a). This was As duodenal DMT-1 mRNA and protein levels are paralleled by an increased incorporation of absorbed radio- surprisingly up regulated with prolonged experimental iron active iron within duodenal ferritin in the experimentally overload, DMT-1 appeared unlikely to be the gate keeper I. Theurl et al. / Journal of Hepatology 43 (2005) 711–719 717

iron overloaded animals which is in accordance with recent data showing that DMT-1 expression is rather affected by luminal iron availability [52], while ferroportin expression responds to systemic iron needs [53] which is also indicative from the negative correlation between duodenal ferroportin mRNA and liver hepcidin mRNA expression levels in iron overloaded mice. Hepcidin interacts with ferroportin mediated iron transfer from the enterocyte as hepcidin induces the internalisation of ferroportin, thereby reducing its iron transport capacity [40]. Recent in vivo data showing that hepcidin injection into rats [53] and mice affect duodenal iron absorption [54] also support this concept. However, when investigating control animals in our study hepcidin mRNA levels were negatively associated with DMT-1 but not with ferroportin mRNA expression which resembles the observations made by Laftah et al. [54] in normal and iron deficient mice. Thus, hepcidin may exert divergent effects on duodenal iron absorption depending on Fig. 5. Effect of secondary iron overload on iron uptake, transfer to the plasma and incorporation of Fe59 into ferritin core complex in the body iron status [55,56]. enterocyte. Control mice and iron loaded mice (five i.p. injections of This leads to the next question of how excess iron, dextran) were orally fed 1.05 mg/ml 59Ferric chloride (specific activity e.g. acquired by multiple transfusions, is handled within O3 mCi/mg Fe) in 0.1 M hydrochloric acid diluted in 100 ml NaCl0.9% non-hemochromatotic individuals. In this context, macro- using a gastric tube. Mice were sacrificed 30 min later. (A) Duodenal iron uptake was defined as the radioactive iron content measured in phages play a decisive role. Following the first i.p. extensively rinsed duodenal enterocytes per mg dry weight. (B) injection the iron content of spleen macrophages and Whereas iron transfer was represented by the radioactive serum iron Kupffer cells increased and IRP-activity, as a marker for content. Serum samples were taken short before sacrificing the mice. the intracellular labile iron pool [27] decreased. With Radioactive content in duodenum and serum were estimated using a progressive iron accumulation we observed a continuous gamma counter. Data are presented as meanGSD of six individuals in each group. *P!0.05 for comparison of iron loaded versus control increase in macrophage ferroportin and ferritin expression mice. (C) For the incorporation assay small duodenal of three control which may be referred to translational induction of these and three iron loaded mice were subjected to cytosolic protein genes by iron [49]. Thus, after initial accumulation iron preparation. Two hundred micrograms of protein of the cell lysates was stored within ferritin while the cytoplasmic were loaded on 8% SDS-polyacrylamide gels and electrophoresed under non-reducing conditions using a Tris/Tricine-electrophoresis concentrations of metabolically available iron decreased buffer. as indicated by a progressive induction of IRP binding affinity. The latter may also referred to the fact that iron is exported from macrophages/Kupffer cells by ferropor- for the control of duodenal iron absorption. Accordingly, an tin. Such a strategy would make sense form the in vivo radioactive iron uptake assay demonstrated that iron immunological point of view since excess iron within uptake into enterocytes is even higher in iron overloaded cells of the reticulo-endothelial system is associated with than in the control animals. Yet, the transfer of the metal an impaired immune response and an increased suscep- from the enterocyte to the plasma was strongly reduced in tibility towards and cancer [4,57,58]. the iron loaded group, which significantly paralleled the Hepatocytes respond to excess iron with an increased reduced ferroportin expression observed with this condition. expression of hepcidin which may then decrease duodenal Therefore, it is suggestive that iron first accumulates in the iron absorption [59–61]. In addition, iron uptake molecules enterocyte and promotes ferritin translation with subsequent in the liver, such as DMT-1 and TfR1 are down-regulated incorporation of iron into the ferritin core [49] (Fig. 5). This with experimental iron overload while ferroportin dynamic process will then results in a reduction of expression is increased [60], however, since whole liver metabolically available iron in the cytoplasm with sub- tissue has been investigated we cannot differentiate, at least sequent activation of IRP binding affinity leading to at mRNA levels, whether these expressional changes are increased expression of DMT-1 and TfR most likely via rather related to hepatocytes or Kupffer cells. IRP mediated stabilisation of the respective mRNAs [50]. In summary, we have demonstrated the dynamic changes This is also reflected by our observation of increased IRP occurring in body iron homeostasis with secondary iron activity and high TfR1 and DMT-1 mRNA expression in overload and enlightened some aspects of the regulatory secondary iron overloaded mice [46,51]. network which controls iron accumulation in this setting Thus, iron taken up from the duodenal lumen is not involving metabolic changes in the liver, spleen and the transferred to the circulation in our model of experimentally duodenum. 718 I. Theurl et al. / Journal of Hepatology 43 (2005) 711–719

Acknowledgements [22] Ludwiczek S, Theurl I, Bahram S, Schumann K, Weiss G. Regulatory networks for the control of body iron homeostasis and their This study was supported by grants from the Austrian dysregulation in HFE mediated hemochromatosis. J Cell Physiol 2005. Research Fund, P-14215 and P-15943. [23] Gunshin H, Allerson CR, Polycarpou-Schwarz M, Rofts A, Rogers JT, Kishi F, et al. Iron-dependent regulation of the divalent metal ion transporter. Fed Eur Biochem Soc Lett 2001;509:309. References [24] Lymboussaki A, Pignatti E, Montosi G, Garuti C, Haile DJ, Pietrangelo A. The role of the iron responsive element in the control of ferroportin1/IREG1/MTP1 gene expression. J Hepatol 2003;39: [1] Hershko C, Link G, Cabantchik I. Pathophysiology of iron overload. 710. Ann N Y Acad Sci 1998;850:191. [25] Tchernitchko D, Bourgeois M, Martin ME, Beaumont C. Expression [2] Kushner JP, Porter JP, Olivieri NF. Secondary iron overload. of the two mRNA isoforms of the iron transporter Nrmap2/DMTI in Hematology (Am Soc Hematol Educ Program) 2001;47. mice and function of the iron responsive element. Biochem J 2002; [3] Gordeuk VR. African iron overload. Semin Hematol 2002;39:263. 363:449. [4] Weiss G. Iron and immunity: a double-edged sword. Eur J Clin Invest [26] Zoller H, Theurl I, Koch R, Kaser A, Weiss G. Mechanisms of iron 2002;32:70. mediated regulation of the duodenal iron transporters divalent metal [5] Halliwell B, Gutteridge JM. Biologically relevant metal ion- transporter 1 and ferroportin 1. Blood Cells Mol Dis 2002;29:488. dependent hydroxyl radical generation. An update. Fed Eur Biochem [27] Frazer DM, Anderson GJ. The orchestration of body iron intake: how Soc Lett 1992;307:108. and where do enterocytes receive their cues? Blood Cells Mol Dis [6] Pietrangelo A. Physiology of iron transport and the hemochromatosis 2003;30:288. gene. Am J Physiol Gastrointest Liver Physiol 2002;282:G403. [28] Feder JN, Gnirke A, Thomas W, Tsuchihashi Z, Ruddy DA, [7] Anderson GJ, Frazer DM, Wilkins SJ, Becker EM, Millard KN, Basava A, et al. A novel MHC class I-like gene is mutated in patients Murphy TL, et al. Relationship between intestinal iron-transporter with hereditary haemochromatosis. Nat Genet 1996;13:399. expression, hepatic hepcidin levels and the control of iron absorption. [29] Davies PS, Zhang AS, Anderson EL, Roy CN, Lampson MA, Biochem Soc Trans 2002;30:724. McGraw TE, et al. Evidence for the interaction of the hereditary [8] Powell LW. Hereditary hemochromatosis and iron overload diseases. haemochromatosis protein, HFE, with the transferrin receptor in J Gastroenterol Hepatol 2002;17:S191. endocytic compartments. Biochem J 2003;373:145. [9] McKie AT, Barrow D, Latunde-Dada GO, Rolfs A, Sager G, [30] Loreal O, Brissot P. Hepcidin: small molecule, large future. Rev Med Mudaly E, et al. An iron-regulated ferric reductase associated with Interne 2003;24:213. the absorption of dietary iron. Science 2001;291:1755. [31] Ganz T. Hepcidin, a key regulator of iron metabolism and mediator of [10] Fleming MD, Trenor 3rd CC, Su MA, Foernzler D, Beier DR, of inflammation. Blood 2003. Dietrich WF, et al. Microcytic anaemia mice have a mutation in [32] Pigeon C, Ilyin G, Courselaud B, Leroyer P, Turlin B, Brissot P, et al. Nramp2, a candidate iron transporter gene. Nat Genet 1997;16:383. A new mouse liver-specific gene, encoding a protein homologous to [11] Gunshin H, Mackenzie B, Berger UV, Gunshin Y, Romero MF, human antimicrobial peptide hepcidin, is overexpressed during iron Boron WF, et al. Cloning and characterization of a mammalian overload. J Biol Chem 2001;276:7811. proton-coupled metal-ion transporter. Nature 1997;388:482. [33] Nicolas G, Bennoun M, Porteu A, Mativet S, Beaumont C, [12] Donovan A, Brownlie A, Zhou Y, Shepard J, Pratt SJ, Moynihan J, Grandchamp B, et al. Severe iron deficiency anemia in transgenic et al. Positional cloning of zebrafish ferroportin1 identifies a mice expressing liver hepcidin. Proc Natl Acad Sci USA 2002;99: conserved vertebrate iron exporter. Nature 2000;403:776. 4596. [13] McKie AT, Marciani P, Rolfs A, Brennan K, Wehr K, Barrow D, et al. [34] Park CH, Valore EV, Waring AJ, Ganz T. Hepcidin, a urinary A novel duodenal iron-regulated transporter, IREG1, implicated in the antimicrobial peptide synthesized in the liver. J Biol Chem 2001;276: basolateral transfer of iron to the circulation. Mol Cell 2000;5:299. 7806. [14] Abboud S, Haile DJ. A novel mammalian iron-regulated protein [35] Weinstein DA, Roy CN, Fleming MD, Loda MF, Wolfsdorf JI, involved in intracellular iron metabolism. J Biol Chem 2000;275: Andrews NC. Inappropriate expression of hepcidin is associated with 19906. iron refractory anemia: implications for the anemia of chronic . [15] Vulpe CD, Kuo YM, Murphy TL, Cowley L, Askwith C, Libina N, Blood 2002;100:3776. et al. Hephaestin, a homologue implicated in intestinal [36] Muckenthaler M, Roy CN, Custodio AO, Minana B, deGraaf J, iron transport, is defective in the sla mouse. Nat Genet 1999;21:195. Montross LK, et al. Regulatory defects in liver and intestine implicate [16] Morgan EH, Oates PS. Mechanisms and regulation of intestinal iron abnormal hepcidin and Cybrd1 expression in mouse hemochroma- absorption. Blood Cells Mol Dis 2002;29:384. tosis. Nat Genet 2003;34:102. [17] Conrad ME, Umbreit JN. Pathways of iron absorption. Blood Cells [37] Nicolas G, Andrews NC, Kahn A, Vaulont S. Hepcidin, a candidate Mol Dis 2002;29:336. modifier of the hemochromatosis phenotype in mice. Blood 2004;103: [18] Parkkila S, Niemela O, Britton RS, Fleming RE, Waheed A, 2841. Bacon BR, et al. Molecular aspects of iron absorption and HFE [38] Roetto A, Papanikolaou G, Politou M, Alberti F, Girelli D, expression. Gastroenterology 2001;121:1489. Christakis J, et al. Mutant antimicrobial peptide hepcidin is associated [19] Chen H, Su T, Attieh ZK, Fox TC, McKie AT, Anderson GJ, et al. with severe juvenile hemochromatosis. Nat Genet 2003;33:21. Systemic regulation of HEPHAESTIN and IREG1 revealed in studies [39] Bridle KR, Frazer DM, Wilkins SJ, Dixon JL, Purdie DM, of genetic and nutritional iron deficiency. Blood 2003. Crawford DH, et al. Disrupted hepcidin regulation in HFE-associated [20] Zoller H, Koch RO, Theurl I, Obrist P, Pietrangelo A, Montosi G, haemochromatosis and the liver as a regulator of body iron et al. Expression of the duodenal iron transporters divalent-metal homoeostasis. Lancet 2003;361:669. transporter 1 and ferroportin 1 in iron deficiency and iron overload. [40] Nemeth E, Tuttle MS, Powelson J, Vaughn MB, Donovan A, Gastroenterology 2001;120:1412. Ward DM, et al. Hepcidin regulates iron efflux by binding to [21] Brissot P, Troadec MB, Loreal O. The clinical relevance of new ferroportin and inducing its internalization. Science 2004. insights in iron transport and metabolism. Curr Hematol Rep 2004;3: [41] Frazer DM, Wilkins SJ, Becker EM, Vulpe CD, McKie AT, 107. Trinder D, et al. Hepcidin expression inversely correlates with I. Theurl et al. / Journal of Hepatology 43 (2005) 711–719 719

the expression of duodenal iron transporters and iron absorption in [51] Schumann K, Moret R, Kunzle H, Kuhn LC. Iron regulatory protein as rats. Gastroenterology 2002;123:835. an endogenous sensor of iron in rat intestinal mucosa. Possible [42] Papanikolaou G, Samuels ME, Ludwig EH, MacDonald ML, implications for the regulation of iron absorption. Eur J Biochem Franchini PL, Dube MP, et al. Mutations in HFE2 cause iron overload 1999;260:362. in 1q-linked juvenile hemochromatosis. Nat Genet 2004; [52] Frazer DM, Wilkins SJ, Becker EM, Murphy TL, Vulpe CD, 36:77. McKie AT, et al. A rapid decrease in the expression of DMT1 and [43] Ludwiczek S, Aigner E, Theurl I, Weiss G. Cytokine-mediated Dcytb but not Ireg1 or hephaestin explains the mucosal block regulation of iron transport in human monocytic cells. Blood 2003; phenomenon of iron absorption. Gut 2003;52:340. 101:4148. [53] Yeh KY, Yeh M, Glass J. Hepcidin regulation of ferroportin 1 [44] Weiss G, Goossen B, Doppler W, Fuchs D, Pantopoulos K, Werner- expression in the liver and intestine of the rat. Am J Physiol Felmayer G, et al. Translational regulation via iron-responsive Gastrointest Liver Physiol 2004;286:G385. elements by the nitric oxide/NO-synthase pathway. Eur Mol Biol [54] Laftah AH, Ramesh B, Simpson RJ, Solanky N, Bahram S, Org J 1993;12:3651. Schumann K, et al. Effect of hepcidin on intestinal iron absorption [45] Leibold EA, Munro HN. Cytoplasmic protein binds in vitro to a highly in mice. Blood 2004;103:3940. conserved sequence in the 50 untranslated region of ferritin heavy- and [55] Kulaksiz H, Gehrke SG, Janetzko A, Rost D, Bruckner T, light-subunit mRNAs. Proc Natl Acad Sci USA 1988;85:2171. Kallinowski B, et al. Pro-hepcidin: expression and cell specific [46] Ludwiczek S, Theurl I, Artner-Dworzak E, Chorney M, Weiss G. localisation in the liver and its regulation in hereditary haemochroma- Duodenal HFE expression and hepcidin levels determine body iron tosis, chronic renal insufficiency, and renal anaemia. Gut 2004;53:735. homeostasis: modulation by genetic diversity and dietary iron [56] Finch C. Regulators of iron balance in humans. Blood 1994;84:1697. availability. J Mol Med 2004;82:373. [57] Gordeuk VR, McLaren CE, MacPhail AP, Deichsel G, Bothwell TH. [47] Perls M. Nachweis von Eisenoxyd in gewissen Pigmentationen. Associations of iron overload in Africa with hepatocellular carcinoma Virchows Archiv fu¨r pathologische Anatomie und klinische Medizin and tuberculosis: Strachan’s 1929 Thesis revisited. Blood 1996;87:3470. 1867;39:42. [58] Weinberg ED. Iron loading and disease surveillance. Emerg Infect Dis [48] Beaumont C, Torti SV, Torti FM, Massover WH. Novel properties of 1999;5:346. L-type polypeptide subunits in mouse ferritin molecules. J Biol Chem [59] Pietrangelo A. Hereditary hemochromatosis—a new look at an old 1996;271:7923. disease. N Engl J Med 2004;350:2383. [49] Arosio P, Levi S. Ferritin, iron homeostasis, and oxidative damage. [60] Ganz T. Hepcidin, a key regulator of iron metabolism and mediator of Free Radic Biol Med 2002;33:457. anemia of inflammation. Blood 2003;102:783. [50] Hentze MW, Muckenthaler MU, Andrews NC. Balancing acts: [61] Zanninelli G, Loreal O, Brissot P, Konijn AM, Slotki IN, Hider RC, molecular control of mammalian iron metabolism. Cell 2004;117: et al. The labile iron pool of hepatocytes in chronic and acute iron 285. overload and chelator-induced iron deprivation. J Hepatol 2002;36:39.